We have arranged for the original English version of State Views and Local Views of Population to be open access. We published this paper some time ago in History and Computing. For families recorded in our Liaoning household registers whose descendants had provided us with their genealogies, we linked the materials across the two sources to assess which of the people recorded in the registers were more or less likely to be recorded in family genealogies.
We found biases in recording in genealogies not anticipated by researchers who have used them for demographic studies. While the tendency of genealogies to omit children who died in infancy and daughters is well-known, we also revealed a tendency for men who survived to adulthood to be less likely to be recorded if they had no offspring, even if they married. Whereas omission of children biases fertility estimates downward by deflating the denominator, omission of adults who had no children biases estimates upward by deflating the numerator.
While the relative importance of these two types of bias will likely vary across genealogies, we believe that the results should caution against simple assumptions that fertility estimates from genealogies are always downward biased and can be corrected by adjusting the number of births to account for infant or child mortality. Here is the link:
https://www.euppublishing.com/doi/abs/10.3366/hac.2002.14.1-2.9